—Dr. Mehmet Sercan Ercan, independent researcher focusing on constitutional theory and comparative constitutional law
Constitution-making in deeply divided societies asks the wrong first question. The real question is not institutional but constitutive: who possesses the authority to constitute the political community in whose name power will be exercised? Constitution-making is usually treated instead as a problem of institutional design – presidentialism or parliamentarism, federalism or unitarism, power-sharing or majoritarian rule. These choices matter, but they all presuppose the resolution of the prior question above, which is however too often left unexamined. The central claim of this post is that, in deeply divided societies, constitutional legitimacy depends not only on how power is institutionally distributed, but on who is included in the exercise of constituent power.
A constitution can distribute power skilfully, accommodate competing interests, and even end an armed conflict, without acquiring the democratic legitimacy needed to endure. Conversely, a constitutional order can remain resilient even while leaving deep disagreements unresolved. Institutional design alone cannot explain this difference.
This post develops that argument through two contrasting cases. Bosnia and Herzegovina shows how a peace settlement can end a war while failing to build an enduring constitutional foundation. Spain shows the opposite trajectory: despite continuing and profound territorial disagreement, an inclusive constituent process produced a framework that has retained its legitimacy as the arena within which that disagreement is contested. Chile’s failed constitutional experiments of 2022-2023 complicate this picture in a different way: both were highly participatory, yet still failed because constituent majorities used the process to entrench their own programmes rather than to build a shared framework.
When Peace Replaces Constituent Power
In Bosnia and Herzegovina, the Dayton Peace Agreement achieved its most important objective: it ended one of Europe’s bloodiest post-war conflicts and prevented a return to large-scale violence. Judged as a peace agreement, Dayton was a significant achievement. Judged as the foundation of a constitutional order, its limitations are far more visible.
The problem is not that Bosnia and Herzegovina adopted power-sharing. Divided societies often need mechanisms that recognise competing identities and prevent domination by a single group, and such mechanisms are compatible with democratic governance when they emerge from a legitimate political agreement. The deeper problem is that Dayton’s mechanisms were never the product of an inclusive constituent process. They emerged from international negotiations aimed at ending a war. Annex 4 became a Constitution without first becoming a social compact.
This is not to deny the constraints under which Dayton was negotiated. Given the military balance, the humanitarian emergency, and the absence of trust among the parties, a broader constituent process was arguably not a realistic option at the time – and peace sometimes requires agreements negotiated among those with the capacity to keep fighting. But this does not change the constitutional point: a peace settlement and a constitution perform different functions and require different sources of legitimacy. The difficulty is not that Dayton ended the war through elite bargaining, but that the same bargain became the permanent constitutional order without a later process capable of supplying democratic legitimacy. Ordinary citizens, civil society, and groups outside the three recognised constituent peoples had no direct role in negotiating the constitutional text, and the Constitution was never submitted to a popular vote.
Bosnia and Herzegovina’s Constitution recognises three “constituent peoples” (Bosniaks, Croats, and Serbs) and structures key institutions, including the three-member Presidency and the upper house of Parliament, around ethnic representation and quotas. By placing ethnic communities at the centre of constitutional representation, Dayton also entrenched a structure in which political authority is permanently organised around ethnic categories, making it difficult for a shared democratic citizenship to develop. This is the “Dayton paradox”: arrangements designed to manage an exceptional post-war moment – ethnic vetoes, guaranteed representation, extensive power-sharing – became permanent features of governance, and the mechanisms that produced negative peace now constrain the development of positive peace. This is part of why reform has proven so difficult. Those who would need to lead it are often the same actors the current structure protects.
Constituent Legitimacy Through Inclusion
Spain’s 1978 Constitution emerged from a transition marked by the same kind of division Bosnia would later face, with competing political identities, unresolved territorial questions, and deep disagreement over the shape of the state. What set the two processes apart was not the institutions each eventually adopted, but who took part in creating them.
The legitimacy of the 1978 Constitution rested on an inclusive constituent process with clear institutional features, not merely a rhetorical commitment to dialogue. The text was negotiated among the major political currents of post-authoritarian Spain – the centrist UCD, the Socialists, the Communists, conservative forces, and Catalan and Basque nationalist parties. Voters then approved it in a 1978 referendum, with 87.87% of votes cast in favour. And the process itself took roughly eighteen months, allowing competing actors to negotiate, revise, and gradually build a shared framework. Inclusiveness, in other words, meant more than different groups sitting at the same table. It meant participation, deliberation, and democratic authorisation.
This is why the Spanish Constitution could survive continuing, even irresolvable, disagreement. It deliberately left the relationship between the unity of the Spanish nation and the recognition of “nationalities and regions” open, using strategic ambiguity to create political space where immediate agreement was impossible. But ambiguity is not a solution in itself – it postpones rather than resolves. The territorial ambiguity of 1978 resurfaced in the fight over Catalonia’s Statute of Autonomy, especially after the Constitutional Court’s 2010 judgment (STC 31/2010, 28 June, FJ 12), showing that deferred constitutional questions do not disappear; they migrate to a different institutional arena, often the courts, exposing them to charges of judicial overreach.
Unlike Dayton, though, the Spanish process supplies a baseline of democratic authorisation that allows these disputes to be contested as political questions rather than as challenges to the constitutional order itself. That is the more general lesson: strategic ambiguity is valuable only when it operates within a Constitution created through a legitimate, inclusive process. It cannot substitute for political agreement indefinitely.
A Chilean Complication
Chile’s failed constitutional processes of 2022 and 2023 test the limits of a purely procedural account of legitimacy. Unlike what happened in Bosnia and Herzegovina, Chile’s drafts were not short on participation – both were produced by elected constituent bodies and submitted to referendum. They failed anyway, largely because the political tendency holding a working majority in each assembly treated the process as an opportunity to entrench its own programme – on pensions, natural resources, the structure of the state – rather than to build a framework the whole society could recognise as its own. The lesson is that inclusion is necessary but not sufficient. A constituent process also needs restraint, a willingness to distinguish foundational commitments from the ordinary business of political majorities. Without it, even a highly participatory process can produce a text that a majority of voters experience as something imposed on them, not something they helped write.
Pluralism as a Constituent Principle
The comparison points to a broader principle. Legal pluralism is usually treated as a constitutional outcome, the recognition of minority rights, linguistic protections, or federal arrangements after a constitution is adopted. But before pluralism can be protected as a value, it has to shape the process through which constitutional authority is created in the first place. Different communities must be recognised as legitimate participants in constituent power, not merely as beneficiaries of the resulting text. Bosnia and Herzegovina shows the danger of the alternative – not recognising diversity, but freezing it, embedding ethnic communities as permanent constitutional subjects in a way that constrains the development of a shared citizenship. Spain shows that a Constitution can acknowledge deep territorial and cultural differences without transforming that difference into competing units of sovereignty, provided the underlying process was inclusive to begin with.
Conclusion
Constitution-making in divided societies continues to start in the wrong place – with institutional architecture, before asking whether the constituent process itself can generate legitimacy. Bosnia and Herzegovina shows that a peace agreement and a Constitution serve different functions, and that mechanisms built to manage conflict can calcify into permanent rigidity once mistaken for a constitutional settlement. Spain shows that a Constitution can remain legitimate despite deep and continuing disagreement, as long as competing groups recognise themselves as its authors. Chile adds the necessary caveat: inclusion alone does not guarantee legitimacy if a process is captured by a majority unwilling to distinguish constitutional commitments from ordinary politics. Institutions – federal or unitary, presidential or parliamentary, power-sharing or majoritarian – remain secondary to a prior question: who has the authority to constitute the political community, and how do they exercise it? Institutional design still matters, but it comes second. Constituent power must first be exercised inclusively, and then with enough restraint to produce a framework that competing groups can recognise as their own.
Suggested citation: Mehmet Sercan Ercan, Beyond Institutional Design: Constituent Power and Legitimacy in Divided Societies, Int’l J. Const. L. Blog, Aug. 23, 2026, at: http://www.iconnectblog.com/beyond-institutional-design-constituent-power-and-legitimacy-in-divided-societies/